Electrocardiographic changes, blood pressure, and working capacity (total work) were recorded during an upright cycle exercise test in 65 children and young adults (mean age 12 years) with valvular or discrete subvalvular aortic stenosis. All patients had cardiac catheterization, but none had surgical treatment. Controls consisted of 164 normal subjects (mean age 15 years). In the patients, the intervals between the exercise test and cardiac catheterization were 1 month or less in 66%, 2 to 12 months in 28%, and 12 to 24 months in 6%. The patients were classified by resting left ventricular to aortic peak systolic pressure gradient into Groups I (gradient less than 30 mm Hg), II (30 to 49 mm Hg), III (50 to 69 mm Hg), and IV (70 mm Hg or greater). The mean frequency and magnitude of exercise-induced S-T depression were greater in the patients than in the control subjects (p less than 0.005) and increased with increasing obstruction in the patients. Mean total work and peak exercise systolic pressure were significantly decreased in the patients with a left ventricular to aortic systolic gradient of 30 mm Hg or greater as compared with the control subjects (p less than 0.03). An exercise profile consisting of S-T depression of 2 mm or more, and markedly decreased total work and systolic blood pressure were characteristic of the 19 patients (Group IV) with a resting left ventricular to aortic systolic gradient of 70 mm Hg or greater. At least 2 or more components of this exercise profile occurred in 11 (24%) of the 46 patients with a resting left ventricular to aortic systolic gradient less than 70 mm Hg. This study demonstrates that exercise testing reveals serious abnormalities in patients otherwise classified as having trivial or moderate obstruction, and that a properly performed exercise study can be done at minimal risk to the patient.